Low-density lipoproteins cause atherosclerotic

cardiovascular disease 1 Evidence from

genetic epidemiologic and clinical studies

A consensus statement from the European

Atherosclerosis Society Consensus Panel

Brian A Ference1 Henry N Ginsberg2 Ian Graham3 Kausik K Ray4 Chris J

Packard5 Eric Bruckert6 Robert A Hegele7 Ronald M Krauss8 Frederick J Raal9

Heribert Schunkert1011 Gerald F Watts12 Jan Boren13 Sergio Fazio14

Jay D Horton1516 Luis Masana17 Stephen J Nicholls18 Boslashrge G Nordestgaard192021

Bart van de Sluis22 Marja-Riitta Taskinen23 Lale Tokgozoglu24 Ulf Landmesser2526

Ulrich Laufs27 Olov Wiklund2829 Jane K Stock30 M John Chapman31dagger and

Alberico L Catapano32dagger

1Division of Translational Research and Clinical Epidemiology Division of Cardiovascular Medicine Wayne State University School of Medicine Detroit MI 48202 USA 2IrvingInstitute for Clinical and Translational Research Columbia University New York NY USA 3Trinity College Dublin Ireland 4Department of Primary Care and Public HealthImperial Centre for Cardiovascular Disease Prevention Imperial College London UK 5College of Medical Veterinary and Life Sciences University of Glasgow Glasgow UK6INSERM UMRS1166 Department of Endocrinology-Metabolism ICAN - Institute of CardioMetabolism and Nutrition AP-HP Hopital de la Pitie Paris France 7Department ofMedicine Robarts Research Institute Schulich School of Medicine and Dentistry University of Western Ontario London Ontario Canada 8Department of AtherosclerosisResearch Childrenrsquos Hospital Oakland Research Institute Oakland CA 94609 USA 9Faculty of Health Sciences University of Witwatersrand Johannesburg South Africa10Deutsches Herzzentrum Munchen Technische Universiteuroat Munchen Munich 80636 Germany 11Deutsches Zentrum fur Herz und Kreislauferkrankungen (DZHK) Partner SiteMunich Heart Alliance Munich 81377 Germany 12Lipid Disorders Clinic Centre for Cardiovascular Medicine Royal Perth Hospital School of Medicine and PharmacologyUniversity of Western Australia Perth Western Australia Australia 13Department of Molecular and Clinical Medicine Wallenberg Laboratory Sahlgrenska University HospitalUniversity of Gothenburg Gothenburg Sweden 14Department of Medicine Center for Preventive Cardiology of the Knight Cardiovascular Institute Oregon Health and ScienceUniversity Portland OR USA 15Department of Molecular Genetics University of Texas Southwestern Medical Center Dallas TX USA 16Department of Internal MedicineUniversity of Texas Southwestern Medical Center Dallas TX USA 17Research Unit of Lipids and Atherosclerosis University Rovira i Virgili C Sant Llorenc 21 Reus 43201Spain 18South Australian Health and Medical Research Institute University of Adelaide Adelaide Australia 19Department of Clinical Biochemistry and The Copenhagen GeneralPopulation Study Herlev and Gentofte Hospital Copenhagen University Hospital Denmark 20Faculty of Health and Medical Sciences University of Copenhagen Denmark21The Copenhagen City Heart Study Frederiksberg Hospital Copenhagen University Hospital Denmark 22Department of Pediatrics Molecular Genetics Section University ofGroningen University Medical Center Groningen Antonius Deusinglaan 1 Groningen AV 9713 The Netherlands 23Helsinki University Central Hospital and Research ProgramsrsquoUnit Diabetes and Obesity Heart and Lung Centre University of Helsinki Helsinki Finland 24Hacettepe University Ankara Turkey 25Department of Cardiology Charite-Universiteuroatsmedizin Berlin (CBF) Hindenburgdamm 30 Berlin 12203 Germany 26Berlin Institute of Health (BIH) Deutsches Zentrum fur Herz-Kreislaufforschung (DZHK)Berlin Germany 27Klinik fur Innere Medizin III Kardiologie Angiologie und Internistische Intensivmedizin Universiteuroatsklinikum des Saarlandes Homburg Saar Germany28Department of Molecular and Clinical Medicine University of Gothenburg Gothenburg Sweden 29Department of Cardiology Sahlgrenska University Hospital GothenburgSweden 30European Atherosclerosis Society Gothenburg Sweden 31INSERM Dyslipidemia and Atherosclerosis Research and University of Pierre and Marie Curie Pitie-Salpetriere University Hospital Paris France and 32Department of Pharmacological and Biomolecular Sciences University of Milan and IRCCS Multimedica Milan Italy

Received 30 November 2016 revised 20 February 2017 editorial decision 1 March 2017 accepted 8 March 2017

The opinions expressed in this article are not necessarily those of the Editors of the European Heart Journal or of the European Society of Cardiology

Corresponding author Telthorn1 313 745 1410 Fax thorn1 313 745 5565 Email bferencemedwayneedudagger These authors contributed equally as senior authors

VC The Author 2017 Published on behalf of the European Society of CardiologyThis is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (httpcreativecommonsorglicensesby-nc40)which permits non-commercial re-use distribution and reproduction in any medium provided the original work is properly cited For commercial re-use please contactjournalspermissionsoupcom

European Heart Journal (2017) 0 1ndash14 CURRENT OPINIONdoi101093eurheartjehx144

Aims To appraise the clinical and genetic evidence that low-density lipoproteins (LDLs) cause atherosclerotic cardiovas-cular disease (ASCVD)

Methodsand results

We assessed whether the association between LDL and ASCVD fulfils the criteria for causality by evaluating thetotality of evidence from genetic studies prospective epidemiologic cohort studies Mendelian randomization stud-ies and randomized trials of LDL-lowering therapies In clinical studies plasma LDL burden is usually estimated bydetermination of plasma LDL cholesterol level (LDL-C) Rare genetic mutations that cause reduced LDL receptorfunction lead to markedly higher LDL-C and a dose-dependent increase in the risk of ASCVD whereas rare var-iants leading to lower LDL-C are associated with a correspondingly lower risk of ASCVD Separate meta-analysesof over 200 prospective cohort studies Mendelian randomization studies and randomized trials including morethan 2 million participants with over 20 million person-years of follow-up and over 150 000 cardiovascular eventsdemonstrate a remarkably consistent dose-dependent log-linear association between the absolute magnitude ofexposure of the vasculature to LDL-C and the risk of ASCVD and this effect appears to increase with increasingduration of exposure to LDL-C Both the naturally randomized genetic studies and the randomized interventiontrials consistently demonstrate that any mechanism of lowering plasma LDL particle concentration should reducethe risk of ASCVD events proportional to the absolute reduction in LDL-C and the cumulative duration of expo-sure to lower LDL-C provided that the achieved reduction in LDL-C is concordant with the reduction in LDL par-ticle number and that there are no competing deleterious off-target effects

Conclusion Consistent evidence from numerous and multiple different types of clinical and genetic studies unequivocally estab-

lishes that LDL causes ASCVD

Keywords Atherosclerosis bull Cardiovascular disease bull Low-density lipoprotein bull Mendelian randomization bull Clinicaltrials bull Statin bull Ezetimibe bull PCSK9 bull Causality bull Recommendations

Introduction

Atherosclerotic cardiovascular disease (ASCVD) and its clinical mani-festations such as myocardial infarction (MI) and ischaemic strokeare the leading cause of morbidity and mortality throughout theworld Multiple exposures have been reported to be associated withan increased risk of cardiovascular events1 The most extensivelystudied of these exposures by far is low-density lipoprotein (LDL)Multiple lines of evidence have established that cholesterol-rich LDLand other apolipoprotein B (apoB)-containing lipoproteins includingvery low-density lipoproteins (VLDL) and their remnants intermedi-ate density lipoproteins (IDL) and lipoprotein(a) [Lp(a)] are directlyimplicated in the development of ASCVD2

Despite this extensive body of evidence however some stillexpress scepticism of the causal nature of the relationship be-tween LDL and the development of ASCVD3 With the availabilityof new highly efficacious LDL-lowering agents and the develop-ment of additional novel lipid-lowering agents with prolongedduration of action there is a need for a consensus as to whetherLDL causes ASCVD in order to inform treatment guidelines andhelp shape regulatory agency guidance for the approval of newmedicines

This Consensus Statement the first of two that evaluatesthe case for LDL causality appraises evidence from genetic epide-miologic and clinical intervention studies The second paper dis-cusses the evidence for LDL causality based on the currentunderstanding of the pathophysiology of ASCVD While our focusis on LDL this does not diminish the importance of the role ofother apoB-containing lipoproteins on the development ofASCVD nor does it exclude potential atherogenic actions of the

individual components of the lipidome and proteome of LDL be-yond cholesterol and apoB

Most publications that question the causal effect of LDL on the de-velopment of ASCVD tend to cite evidence from individual studiesor a small group of highly selected studies often without a quantita-tive synthesis of the presented evidence4 Therefore to avoid thistype of selection bias we have based our conclusions on the totalityof evidence from separate meta-analyses of genetic studies prospect-ive epidemiologic studies Mendelian randomization studies andrandomized clinical trials This evidence base includes over 200 stud-ies involving over 2 million participants with over 20 million person-years of follow-up and more than 150 000 cardiovascular eventsTogether these studies provide remarkably consistent and unequivo-cal evidence that LDL causes ASCVD as summarized in Table 1

Pathophysiology of atherosclerosis

The key events in the initiation of ASCVD are the retention and accu-mulation of cholesterol-rich apoB-containing lipoproteins within thearterial intima at sites of predilection for plaque formation5ndash9

Notably LDL and other apoB-containing lipoproteins lt70 nm indiameter (including VLDL their remnants IDL and Lp(a)) efficientlyenter and exit the arterial intima910 At what is now consideredphysiologically relevant levels of LDL cholesterol [LDL-C 05ndash10 mmolL (20ndash40 mgdL) typical of newborns11 and a wide range ofmammalian species12] the probability of LDL particle retention andthe risk of developing atherosclerosis is low6 However as the con-centration of LDL-C increases above this level the probability of in-timal retention of LDL leading to the initiation and progressive

2 BA Ference et al

development of atherosclerotic plaque increases in a dose-dependent manner2 This topic will be discussed in detail in the se-cond Consensus Statement on this topic

Cholesterol LDL and LDL-C

The terms lsquocholesterolrsquo lsquoLDLrsquo and lsquoLDL cholesterol (LDL-C)rsquo are fre-quently conflated or used interchangeably potentially leading to con-fusion Cholesterol is an essential component of cell membranes anda precursor of bile acids and steroid hormones Importantly choles-terol of both exogenous and endogenous origin is transported toperipheral cells largely by the apoB-containing lipoproteins in plasmaIn most people LDL particles constitute 90 of circulating apoB-containing lipoproteins in fasting blood (Figure 1) However in clinicalpractice the plasma LDL level is generally not measured directly butinstead is estimated from its cholesterol concentrationmdashLDL-Cmdashameasure of the total amount of cholesterol contained in LDL par-ticles As a result calculated plasma LDL-C has become the focus forassessing cardiovascular risk and for evaluating therapeutic benefit inrandomized clinical trials

Under most conditions LDL-C concentration and LDL particlenumber are highly correlated and therefore plasma LDL-C is a good

surrogate for LDL particle concentration However in certain condi-tions (eg the metabolic syndrome diabetes and hypertriglyceridae-mia) plasma LDL-C and LDL particle concentration can becomediscordant as a result of the predominance of small densecholesterol-poor LDL and therefore plasma LDL-C may not accur-ately reflect LDL particle concentration or its effect on cardiovascularrisk Under these conditions direct measurement of LDL particlenumber or apoB concentration (recognizing that each LDL particlecontains a single apoB molecule) may more accurately reflect thecausal effect of LDL on ASCVD In this Consensus Statement we as-sess the evidence that LDL causes ASCVD by critically appraising theclinical evidence with the understanding that the majority of clinicalstudies have used LDL-C as an estimate of LDL concentration

Evidence from inherited disordersof lipid metabolism

Familial hypercholesterolaemia (FH) is an autosomal co-dominantdisorder that usually results from a loss-of-function (LOF) mutationin the LDL receptor (LDLR) gene or less commonly from LOF muta-tions in the APOB gene that reduce the binding of apoB-containinglipoproteins to the LDL receptor or a gain-of-function (GOF)

Table 1 Criteria for causality low-density lipoprotein (LDL) and atherosclerotic cardiovascular disease (ASCVD)

Criterion (modified

from reference5)

Evidence

grade

Summary of the evidence (references)

1 Plausibility 1 LDL and other apolipoprotein (apo) B-containing lipoproteins (very low-density lipoprotein their remnants

intermediate-density lipoprotein and lipoprotein(a)) are directly implicated in the initiation and progression

of ASCVD experimentally induced elevations in plasma LDL and other apoB-containing lipoproteins lead to

atherosclerosis in all mammalian species studied25ndash12

2 Strength 1 Monogenic and polygenic-mediated lifelong elevations in LDL lead to markedly higher lifetime risk13ndash2027ndash314043

3 Biological gradient 1 Monogenic lipid disorders prospective cohort studies Mendelian randomization studies and randomized inter-

vention trials uniformly demonstrate a dose-dependent log-linear association between the absolute magni-

tude of exposure to LDL and risk of ASCVD13ndash2227ndash3638ndash4042ndash47

4 Temporal sequence 1 Monogenic lipid disorders and Mendelian randomization studies demonstrate that exposure to elevated LDL

precedes the onset of ASCVD13ndash2027ndash314043

5 Specificity 1 Mendelian randomization studies and randomized intervention trials both provide unconfounded randomized

evidence that LDL is associated with ASCVD independent of other risk factors2831ndash334043

6 Consistency 1 Over 200 studies involving more than 2 million participants with over 20 million person-years of follow-up and

more than 150 000 cardiovascular events consistently demonstrate a dose-dependent log-linear association

between the absolute magnitude of exposure to LDL and risk of ASCVD13ndash2227ndash3638ndash4042ndash47

7 Coherence 1 Monogenic lipid disorders prospective cohort studies Mendelian randomization studies and randomized inter-

vention trials all show a dose-dependent log-linear association between the absolute magnitude of exposure

to LDL and risk of ASCVD15ndash1821222830ndash323536434447

8 Reduction in risk with

intervention

1 More than 30 randomized trials involving over 200 000 participants and 30 000 ASCVD events evaluating thera-

pies specifically designed to lower LDL (including statins ezetimibe and PCSK9 inhibitors) consistently dem-

onstrate that reducing LDL cholesterol (LDL-C) reduces the risk of ASCVD events proportional to the

absolute reduction in LDL-C32ndash3438394245ndash47

Criteria are graded by a modification of the quality criteria adopted by the European Society of Cardiology systemFor reference see httpwwwescardioorgGuidelines-amp-EducationClinical-Practice-GuidelinesGuidelinesdevelopmentWriting-ESC-Guidelines (31 January 2017)These are defined as followsClass 1 Evidence andor general agreement that the criterion for causality is fulfilledClass 2 Conflicting evidence andor a divergence of opinion about whether the criterion indicated causalityClass 3 Evidence or general agreement that the criterion for causality is not fulfilled

Low-density lipoproteins cause atherosclerotic cardiovascular disease 3

mutation in the PCSK9 gene Regardless of the underlying genetic de-fect FH is characterized by markedly elevated levels of LDL-C andpremature atherosclerosis particularly coronary artery disease1314

In the most common form of FH a mutation in the LDLR genecauses decreased LDL receptor function leading to a markedlyincreased concentration of circulating LDL particles and the choles-terol carried by those particles (ie LDL-C) Heterozygous FH(HeFH) affects between 1200 and 1300 people worldwide1516 andwhen untreated is characterized by LDL-C levels in the range of 45ndash12 mmolL and a marked increase in risk of ASCVD13 HomozygousFH (HoFH) is a much rarer condition with an extreme phenotypecharacterized by untreated plasma LDL-C levels often in excess of13 mmolL from birth and almost universal development of ASCVDin childhood or early adolescence14 Although the phenotypic ex-pression of FH is variable the extent of atherosclerosis and the riskof cardiovascular events in both HeFH and HoFH is proportional toboth the absolute magnitude and the duration of exposure to ele-vated LDL-C levels151718

Within any affected family each child has an equal and random 50probability of inheriting a mutation that causes FH The fact that siblingswho inherit an FH mutation have markedly elevated plasma LDL-C lev-els and a corresponding dose-dependent markedly elevated lifetimerisk of ASCVD as compared to their unaffected siblings providespowerful evidence that LDL causes ASCVD19 This conclusion is sub-stantially strengthened by the complementary observations that GOFmutations in PCSK9 result in a markedly elevated LDL-C concentrationand a corresponding markedly elevated risk of ASCVD while LOF mu-tations in PCSK9 result in a lower LDL-C concentration and a corres-ponding markedly lower lifetime risk of ASCVD20

Evidence from prospectiveepidemiologic studies

Several large meta-analyses of prospective observational epidemio-logic studies using individual participant data have consistently re-ported a continuous log-linear association between the absolutemagnitude of exposure to plasma LDL-C levels and the risk ofASCVD

The Emerging Risk Factors Collaboration (ERFC) reported the re-sults of a meta-analysis of individual participant data from 302 430persons without prevalent vascular disease at the time of enrolmentin 68 prospective studies during which 8857 non-fatal MI and 928coronary heart disease (CHD) deaths accrued over 279 millionperson-years of follow-up21 In these studies plasma LDL-C concen-tration was log-linearly associated with increased risk of non-fatal MIor CHD death Although the authors reported the association be-tween non-HDL-C concentration and risk of CHD in the primaryanalysis all studies included in this meta-analysis measured total chol-esterol high-density lipoprotein cholesterol (HDL-C) and triglycer-ides and reported the calculated LDL-C concentration as estimatedby the Friedewald equation The ERFC authors point out that any re-gression model that includes terms for non-HDL-C HDL-C and tri-glycerides is a simple mathematical rearrangement of a model thatincludes terms for calculated LDL-C HDL-C and triglyceridesTherefore in the ERFC analysis the effect of LDL-C is exactly equalto the effect of non-HDL-C on the risk of CHD by definition in theanalysis The authors confirmed this fact by demonstrating that in asubsample of 8 studies involving 44 234 individuals the effect of dir-ectly measured LDL-C on the risk of CHD was nearly identical to theeffect of non-HDL-C (and calculated LDL-C) per millimole per litre

Similarly the Prospective Studies Collaboration reported a meta-analysis of individual participant data on 892 337 persons withoutcardiovascular disease at baseline who had been enrolled in 61 pro-spective cohort studies during which 33 744 ischaemic heart diseasedeaths accrued over nearly 12 million person-years of follow-up Thismeta-analysis reported a strong graded log-linear association be-tween total plasma cholesterol and the risk of ischaemic heart diseasemortality22 Importantly in a subsample of 153 798 participants forwhom HDL-C measurements were available the effect of non-HDL-C on the risk of ischaemic heart disease mortality was nearly identicalto the effect of total cholesterol per millimole per litre

Together these meta-analyses of prospective epidemiologic cohortstudies provide coherent and consistent evidence that plasma LDL-Cconcentration is strongly and log-linearly associated with a dose-dependent increase in the risk of incident ASCVD events (Figure 2)

Evidence from Mendelianrandomization studies

Although the association between LDL-C and the risk of ASCVD isstrong graded and reproducible in meta-analyses of prospective co-hort studies these studies are not randomized and are therefore un-avoidably vulnerable to confounding reverse causation and otherforms of bias Mendelian randomization studies introduce a random-ization scheme into an observational study specifically to assess

Figure 1 Relative concentration of apolipoprotein B (ApoB) con-tained in circulating lipoproteins in normolipidaemic individualsApoB content was calculated in nanomoles per litre using 500 000as the defined molecular mass [ie low-density lipoprotein (LDL)100 mgdL or 2000 nmolL very low-density lipoprotein (VLDL)5 mgdL or 100 nmolL intermediate density lipoprotein (IDL) rem-nants 5 mgdL or 100 nmolL and lipoprotein(a) 10 nmoll] Basedon population median

4 BA Ference et al

whether an observed association between an exposure and an out-come is likely to be causal23

Numerous variants in multiple genes have been reported to beassociated with lower LDL-C levels2425 Each of these variants is in-herited approximately randomly at the time of conception in a pro-cess sometimes referred to as Mendelian randomization Thereforeinheriting an LDL-C lowering allele in one of these genes is analogousto being randomly allocated to treatment with an LDL-C-loweringtherapy while inheriting the other allele is analogous to being ran-domly allocated to lsquousual carersquo If the variant under study is associated

solely with LDL-C and not with other lipid or non-lipid pleiotropic ef-fects and if allocation is indeed random then comparing the risk ofASCVD among persons with and without such a variant should pro-vide an unconfounded estimate of the causal effect of lower LDL-Clevels on the risk of ASCVD in a manner analogous to a long-termrandomized trial26

Mendelian randomization studies have consistently demonstratedthat variants in over 50 genes that are associated with lower LDL-Clevels (but not with other potential predictors or intermediates forASCVD) are also associated with a correspondingly lower risk of

Figure 2 Log-linear association per unit change in low-density lipoprotein cholesterol (LDL-C) and the risk of cardiovascular disease as reportedin meta-analyses of Mendelian randomization studies prospective epidemiologic cohort studies and randomized trials The increasingly steeper slopeof the log-linear association with increasing length of follow-up time implies that LDL-C has both a causal and a cumulative effect on the risk of cardio-vascular disease The proportional risk reduction (y axis) is calculated as 1-relative risk (as estimated by the odds ratio in Mendelian randomizationstudies or the hazard ration in the prospective epidemiologic studies and randomized trials) on the log scale then exponentiated and converted to apercentage The included meta-analyses were identified from (i) MEDLINE and EMBASE using the search terms meta-analysis LDL and lsquocardiovascu-lar or coronaryrsquo (ii) the reference lists of the identified meta-analyses (iii) public data from GWAS consortia and (iv) by discussion with members ofthe EAS Consensus Panel We included the most updated meta-analyses available giving preference to meta-analyses that used individual participantdata Trial acronyms AFTexCAPS Air ForceTexas Coronary Atherosclerosis Prevention Study ALERT Assessment of LEscol in RenalTransplantation ALLHAT-LLT Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial Lipid Lowering Trial ALLIANCEAggressive Lipid-Lowering Initiation Abates New Cardiac Events ASPEN Atorvastatin Study for Prevention of Coronary Heart Disease Endpoints innon-insulin-dependent diabetes mellitus ASCOT LLA Anglo Scandinavian Cardiac Outcomes Trial Lipid Lowering Arm AURORA A Study toEvaluate the Use of Rosuvastatin in Subjects on Regular Hemodialysis An Assessment of Survival and Cardiovascular Events CARE Cholesterol andRecurrent Events CARDS Collaborative Atorvastatin Diabetes Study CHGN Community Health Global Network 4D Deutsche Diabetes DialyseStudies ERFC Emerging Risk Factors Collaboration GISSI Gruppo Italiano per lo Studio della Sopravvivenza nellrsquoInfarto Miocardico HOPE HeartOutcomes Prevention Evaluation Study HPS Heart Protection Study IDEAL Incremental Decrease in End Points Through Aggressive LipidLowering IMPROVE-IT Examining Outcomes in Subjects With Acute Coronary Syndrome Vytorin (EzetimibeSimvastatin) vs Simvastatin JUPITERJustification for the Use of Statins in Primary Prevention An Intervention Trial Evaluating Rosuvastatin trial LIPID Long-Term Intervention withPravastatin in Ischemic Disease LIPS Lescol Intervention Prevention Study MEGA Management of Elevated Cholesterol in the Primary PreventionGroup of Adult Japanese POST-CABG Post Coronary Artery Bypass Graft PROSPER Pravastatin in elderly individuals at risk of vascular diseasePROVE-IT Pravastatin or Atorvastatin Evaluation and Infection Therapy SHARP Study of Heart and Renal Protection TNT Treating to NewTargets WOSCOPS West of Scotland Coronary Prevention Study

Low-density lipoproteins cause atherosclerotic cardiovascular disease 5

CHD2027ndash30 thus providing powerful evidence that LDL is causallyassociated with the risk of CHD Indeed when the effect of eachLDL-C variant is plotted against its effect on CHD there is a continu-ous dose-dependent and log-linear causal association between themagnitude of the absolute change in LDL-C level and the lifetime riskof CHD (Figure 2)2830

Furthermore when adjusted for a standard decrement in LDL-C each of the genetic variants associated with LDL-C has a re-markably similar effect on the risk of CHD per unit lower LDL-Cincluding variants in the genes that encode the targets of pharma-cological agents commonly used to lower LDL-C [ie 3-hydroxy-3-methyl-glutaryl-coenzyme A reductase (HMGCR) the target ofstatins Niemann-Pick C1-like 1 (NPC1L1) the target ofezetimibe and proprotein convertase subtilisinkexin type 9

(PCSK9) the target of the monoclonal antibodies alirocumab andevolocumab see Figure 3] with no evidence of any heterogeneityof effect (I2 = 0)283031 This observation strongly implies thatthe causal effect of these variants on the risk of CHD is mediatedessentially entirely through LDL because it would be implausiblethat variants in numerous different genes involving multiple dis-tinct biological pathways by which LDL is lowered would eachhave directionally concordant and quantitatively similar pleio-tropic effects on the risk of ASCVD

Taken together meta-analyses of Mendelian randomization studiesinvolving more than 300 000 participants and 80 000 CHD cases pro-vide compelling evidence that LDL is causally associated with the riskof ASCVD and that the causal effect of LDL on ASCVD is largely in-dependent of the mechanism by which LDL is lsquoloweredrsquo

Figure 3 Effect of exposure to lower low-density lipoprotein cholesterol (LDL-C) by mechanism of LDL-C lowering Panel A shows the effect ofgenetic variants or genetic scores combining multiple variants in the genes that encode for the targets of currently available LDL-C-lowering thera-pies adjusted for a standard decrement of 035 mmolL lower LDL-C in comparison with the effect of lower LDL-C mediated by variants in the LDLreceptor gene Panel B shows the effect of currently available therapies that act to primarily lower LDL-C through the LDL receptor pathway ad-justed per millimole per litre lower LDL-C Both the naturally randomized genetic data in Panel A and the data from randomized trials in Panel B sug-gest that the effect of LDL-C on the risk of cardiovascular events is approximately the same per unit change in LDL-C for any mechanism that lowersLDL-C via up-regulation of the LDL receptor where the change in LDL-C (which is used in clinical medicine to estimate the change in LDL particleconcentration) is likely to be concordant with changes in LDL particle concentration

6 BA Ference et al

Evidence from randomizedcontrolled trials

Perhaps the most compelling clinical evidence for causality is pro-vided by randomized clinical trials evaluating the effect of therapiesthat reduce LDL-C on the risk of cardiovascular events (For refer-ence Figure 4 shows the principal sites of action of therapeutic inter-ventions that act mainly to lower LDL) It is important to notehowever that the interpretation of any given individual trial can besubstantially affected by its design In general studies that are under-powered due to small sample sizes or few accrued events that donot produce a substantial difference in LDL-C concentration be-tween treatment arms and those with short-term follow-up (2 yearsor less) are more likely to produce results that do not show statistic-ally significant differences Furthermore some therapies that lowerLDL-C (eg oestrogen) also have adverse effects that increase therisk of ASCVD which can attenuate or erase the clinical benefit oflowering LDL-C Over-interpretation of these individual trials maylead to biased conclusions For that reason we focus on the totalityof the evidence derived from meta-analyses of randomized trials

In a meta-analysis of individual-participant data from 26 statin trialsincluding almost 170 000 individuals treatment with a statin was asso-ciated with a log-linear 22 proportional reduction in the risk ofmajor cardiovascular events per millimole per litre reduction in LDL-C over a median of 5 years of treatment32 The effect was some-what less during the first year of treatment followed by a consistent22-24 proportional reduction in cardiovascular events per milli-mole per litre reduction in LDL-C during each subsequent year of

treatment33 The magnitude of this effect was independent of base-line LDL-C level similar among persons with and without pre-existing cardiovascular disease at baseline and remarkably consistentin all subgroups studied3233 This meta-analysis therefore providespowerful evidence that reducing plasma LDL-C levels by inhibitingHMG-CoA reductase with a statin leads to dose-dependent reduc-tion in the risk of major cardiovascular events that is proportional tothe absolute magnitude of the reduction in LDL-C (Figure 2)

Notably in the statin trials the absolute yearly event rate observedin each randomized treatment arm was strongly and linearly associ-ated with the absolute achieved LDL-C level (Figure 5A)34 In thesestudies LDL-C and apoB concentrations had very similar effects onthe risk of cardiovascular events per millimole per litre thus confirm-ing that LDL-C is a satisfactory surrogate for LDL particle numberunder most circumstances Furthermore intravascular ultrasoundstudies of coronary atherosclerosis involving statin-treated patientshave consistently demonstrated that progression of coronary athero-sclerotic plaque volume can be substantially arrested at achievedLDL-C levels of18 mmolL (70 mgdL) (Figure 5B)3536

Importantly other therapies that reduce LDL-C without inhibitingHMG-CoA reductase have also been shown to reduce the risk ofcardiovascular events in large randomized cardiovascular outcometrials Ezetimibe inhibits intestinal absorption of cholesterol by bind-ing to the NPC1L1 transporter protein37 Among the 18 144 personswith acute coronary syndrome enrolled in the IMPROVE-IT trialtreatment with ezetimibe plus a statin vs statin monotherapyreduced LDL-C levels by 040 mmolL over the trial duration andwas associated with a 65 proportional reduction in major cardio-vascular events38 Supportive evidence is available from the SHARP

Figure 4 Schematic figure showing that all therapies that act predominantly to lower low-density lipoprotein (LDL) act via the LDL receptor path-way to up-regulate LDL receptors and thus increase LDL clearance

Low-density lipoproteins cause atherosclerotic cardiovascular disease 7

Figure 5 Linear association between achieved low-density lipoprotein cholesterol (LDL-C) level and absolute coronary heart disease (CHD) eventrate or progression of atherosclerosis Panel A shows absolute cardiovascular event rates in randomized statin trials and Panel B shows progression ofatherosclerosis as measured by intravascular ultrasound In Panel A achieved LDL-C in primary prevention trials and secondary prevention trials in stableCHD patients was related to the end point of CHD events (fatal plus non-fatal myocardial infarction sudden CHD death) proportioned to 5 yearsassuming linear rates with time Trendlines for primary and secondary prevention associations are virtually superimposable Key p placebo a activetreatment arm except for IDEAL where s simvastatin and a atorvastatin and HOPE-3 where r rosuvastatin and TNT where reference is made toatorvastatin 10 and 80 mg dose Trial acronyms AFCAPS Air Force Coronary Atherosclerosis Prevention Study ASCOT Anglo Scandinavian CardiacOutcomes Trial ASTEROID A Study To Evaluate the Effect of Rosuvastatin on Intravascular Ultrasound-Derived Coronary Atheroma Burden CARECholesterol and Recurrent Events CAMELOT Comparison of Amlodipine vs Enalapril to Limit Occurrence of Thrombosis HOPE Heart OutcomesPrevention Evaluation Study HPS Heart Protection Study IDEAL Incremental Decrease in End Points Through Aggressive Lipid LoweringILLUSTRATE Investigation of Lipid Level Management Using Coronary Ultrasound To Assess Reduction of Atherosclerosis by CETP Inhibition andHDL Elevation JUPITER Justification for the Use of Statins in Primary Prevention An Intervention Trial Evaluating Rosuvastatin trial LIPID Long-TermIntervention with Pravastatin in Ischemic Disease PRECISE IVUS Plaque REgression with Cholesterol absorption Inhibitor or Synthesis inhibitorEvaluated by IntraVascular UltraSound PROSPER Pravastatin in elderly individuals at risk of vascular disease REVERSAL Reversal of AtherosclerosisWith Aggressive Lipid Lowering 4S Scandinavian Simvastatin Survival Study SATURN Study of Coronary Atheroma by Intravascular Ultrasound Effectof Rosuvastatin vs Atorvastatin STRADIVARIUS Strategy To Reduce Atherosclerosis Development InVolving Administration of RimonabantmdashtheIntravascular Ultrasound Study TNT Treating to New Targets WOSCOPS West Of Scotland Coronary Prevention Study

8 BA Ference et al

trial in 9270 persons with chronic kidney disease which comparedtreatment with the combination of simvastatin and ezetimibe withplacebo Combination treatment reduced LDL-C levels by085 mmolL (33 mgdL) and reduced the incidence of major cardio-vascular events by 17 relative to placebo39 In both trials the magni-tude of the observed proportional risk reduction was consistent withthe clinical benefit that would be expected from similar absolute re-ductions in LDL-C during treatment with statin monotherapy Thesedata suggest that statins and ezetimibe have therapeutically equivalenteffects on the risk of cardiovascular events per unit lower LDL-C(Figure 3B) These findings also agree closely with the naturallyrandomized genetic data that have demonstrated that variants in theHMGCR and NPC1L1 genes have biologically equivalent effects on therisk of cardiovascular disease per unit lower LDL-C3140 (Figure 3A)

In addition among 27564 participants with cardiovascular diseaseenrolled in the recently reported FOURIER trial treatment with thePCSK9 monoclonal antibody evolocumab added to a statin reducedLDL-C by 14 mmolL (534 mgdl) to a mean LDL-C level of078 mmolL (30 mgdl) and reduced the incidence of cardiovasculardeath MI or stroke by 20 during a median of 22 years of follow-up4142 When analysed by each year of treatment evolocumab hadnearly identical effects on the risk of multiple different cardiovascularendpoints per millimole per litre reduction in LDL-C compared totreatment with a statin as reported by the Cholesterol TreatmentTrialists Collaboration31 The results of this trial are remarkably con-sistent with the results of a recent Mendelian randomization studythat showed that variants in the PCSK9 and HMGCR genes have nearlyidentical effects on the risk of cardiovascular events per unit lowerLDL-C and therefore accurately predicted that PCSK9 inhibitorswould reduce the risk of cardiovascular events by approximately thesame amount as statins per unit reduction in LDL-C43 The results ofthe FOURIER trial are also consistent with the results of the recentGLAGOV trial in which treatment with evolocumab added to a statinreduced LDL-C levels by 145 mmolL and induced plaque regressionthat appeared to be directly proportional to the achieved absoluteLDL-C level even at LDL-C levels as low as 095 mmolL (366 mgdl)44 Taken together the results of these studies strongly suggestthat PCSK9 inhibitors and statins have biologically equivalent effectson the risk of cardiovascular events per unit change in LDL-C

Furthermore in the Lipid Research Clinics Coronary PrimaryPrevention Trial treatment with the bile acid sequestrant cholestyr-amine reduced LDL-C by 07 mmolL and reduced the relative risk ofcardiovascular death or MI by 1945 Similarly in the Program onSurgical Control of Hyperlipidemia (POSCH) study ileal bypass sur-gery reduced LDL-C by 185 mmolL and reduced the relative risk ofcardiovascular death or MI by 3546 In both trials the observed riskreduction was consistent with the proportional risk reduction perunit lower LDL-C observed during treatment with a statin or withcombined statin and ezetimibe therapy (Figure 2) Indeed in a recentmeta-analysis of 49 studies including 312 175 participants and 39 645major vascular events that compared the effects of 9 different typesof lipid-lowering therapies nearly all of these therapeutic approachesto lowering LDL-C were associated with a consistent 20ndash25 pro-portional reduction in vascular events per millimole per litre reduc-tion in LDL-C47 These findings are in line with those from theMendelian randomization studies and demonstrate that the clinical

benefit of lowering LDL-C appears to be independent of the mechan-ism by which LDL-C is lowered (Figure 3)

The one notable exception to this finding is the effect of the cho-lesteryl ester transfer protein (CETP) inhibitors In the recently re-ported ACCELERATE Trial the CETP inhibitor evacetrapib plus astatin reduced LDL-C by 075 mmolL compared with statin mono-therapy but did not reduce the risk of cardiovascular events48 Todate we lack a mechanistic explanation for this finding although dele-terious effects on CHD due to a 15 mmHg mean increase in systolicblood pressure andor possible dysfunctional HDL phenotypes onevacetrapib may constitute contributory factors In addition the LDLparticle-lowering effect of CETP inhibitors appears to become atte-nuated when a CETP inhibitor is added to a statin thus leading to dis-cordance between the measured LDL-C and apoB reductions48 Theongoing 30 000-person REVEAL trial with the CETP inhibitor anace-trapib will provide more definitive evidence of the effect of adding aCETP inhibitor to a statin on the risk of cardiovascular events

Criteria for causality

A critical appraisal of the evidence discussed in this review demon-strates that the association between plasma LDL-C concentrationand the risk of ASCVD satisfies all the criteria for causality (Table 1)49

Indeed the prospective epidemiologic studies Mendelian randomiza-tion studies and randomized intervention trials all demonstrate a re-markably consistent dose-dependent log-linear association betweenthe absolute magnitude of exposure to LDL-C and the risk ofASCVD and together demonstrate that the effect of LDL-C on therisk of ASCVD increases with increasing duration of exposure (Figure2) This concordance between multiple lines of evidence most not-ably the remarkable concordance between the unbiased naturallyrandomized genetic data and the results of numerous randomizedintervention trials using multiple different agents to reduce LDL-Cprovides overwhelming clinical evidence that LDL causes ASCVD andthat lowering LDL reduces the risk of cardiovascular events

Evidence for the cumulative effectof exposure to LDL on ASCVD

Mendelian randomization studies suggest that long-term exposure tolower LDL-C is associated with up to a three-fold greater proportionalreduction in the risk of cardiovascular disease per unit reduction in LDL-C when compared with shorter term treatment with a statin startedlater in life after atherosclerosis has developed28 This finding suggeststhat the causal effect of LDL on the risk of ASCVD is determined byboth the absolute magnitude and the cumulative duration of exposureto LDL-C This finding is consistent with increasing benefit observedover time in the long-term follow-up of the WOSCOPS trial50

Because the effect of LDL-C on the risk of ASCVD appears to beboth causal and cumulative over time lowering the plasma LDL-Clevel earlier than is currently recommended may result in propor-tionally greater reductions in the lifetime risk of ASCVD than whichis estimated in short-term randomized trials Integrating the availableevidence from Mendelian randomization studies and randomized tri-als suggests that each millimole per litre reduction in LDL-C reduces

Low-density lipoproteins cause atherosclerotic cardiovascular disease 9

trial in 9270 persons with chronic kidney disease which comparedtreatment with the combination of simvastatin and ezetimibe withplacebo Combination treatment reduced LDL-C levels by085 mmolL (33 mgdL) and reduced the incidence of major cardio-vascular events by 17 relative to placebo39 In both trials the magni-tude of the observed proportional risk reduction was consistent withthe clinical benefit that would be expected from similar absolute re-ductions in LDL-C during treatment with statin monotherapy Thesedata suggest that statins and ezetimibe have therapeutically equivalenteffects on the risk of cardiovascular events per unit lower LDL-C(Figure 3B) These findings also agree closely with the naturallyrandomized genetic data that have demonstrated that variants in theHMGCR and NPC1L1 genes have biologically equivalent effects on therisk of cardiovascular disease per unit lower LDL-C3140 (Figure 3A)

In addition among 27564 participants with cardiovascular diseaseenrolled in the recently reported FOURIER trial treatment with thePCSK9 monoclonal antibody evolocumab added to a statin reducedLDL-C by 14 mmolL (534 mgdl) to a mean LDL-C level of078 mmolL (30 mgdl) and reduced the incidence of cardiovasculardeath MI or stroke by 20 during a median of 22 years of follow-up4142 When analysed by each year of treatment evolocumab hadnearly identical effects on the risk of multiple different cardiovascularendpoints per millimole per litre reduction in LDL-C compared totreatment with a statin as reported by the Cholesterol TreatmentTrialists Collaboration31 The results of this trial are remarkably con-sistent with the results of a recent Mendelian randomization studythat showed that variants in the PCSK9 and HMGCR genes have nearlyidentical effects on the risk of cardiovascular events per unit lowerLDL-C and therefore accurately predicted that PCSK9 inhibitorswould reduce the risk of cardiovascular events by approximately thesame amount as statins per unit reduction in LDL-C43 The results ofthe FOURIER trial are also consistent with the results of the recentGLAGOV trial in which treatment with evolocumab added to a statinreduced LDL-C levels by 145 mmolL and induced plaque regressionthat appeared to be directly proportional to the achieved absoluteLDL-C level even at LDL-C levels as low as 095 mmolL (366 mgdl)44 Taken together the results of these studies strongly suggestthat PCSK9 inhibitors and statins have biologically equivalent effectson the risk of cardiovascular events per unit change in LDL-C

Furthermore in the Lipid Research Clinics Coronary PrimaryPrevention Trial treatment with the bile acid sequestrant cholestyr-amine reduced LDL-C by 07 mmolL and reduced the relative risk ofcardiovascular death or MI by 1945 Similarly in the Program onSurgical Control of Hyperlipidemia (POSCH) study ileal bypass sur-gery reduced LDL-C by 185 mmolL and reduced the relative risk ofcardiovascular death or MI by 3546 In both trials the observed riskreduction was consistent with the proportional risk reduction perunit lower LDL-C observed during treatment with a statin or withcombined statin and ezetimibe therapy (Figure 2) Indeed in a recentmeta-analysis of 49 studies including 312 175 participants and 39 645major vascular events that compared the effects of 9 different typesof lipid-lowering therapies nearly all of these therapeutic approachesto lowering LDL-C were associated with a consistent 20ndash25 pro-portional reduction in vascular events per millimole per litre reduc-tion in LDL-C47 These findings are in line with those from theMendelian randomization studies and demonstrate that the clinical

benefit of lowering LDL-C appears to be independent of the mechan-ism by which LDL-C is lowered (Figure 3)

The one notable exception to this finding is the effect of the cho-lesteryl ester transfer protein (CETP) inhibitors In the recently re-ported ACCELERATE Trial the CETP inhibitor evacetrapib plus astatin reduced LDL-C by 075 mmolL compared with statin mono-therapy but did not reduce the risk of cardiovascular events48 Todate we lack a mechanistic explanation for this finding although dele-terious effects on CHD due to a 15 mmHg mean increase in systolicblood pressure andor possible dysfunctional HDL phenotypes onevacetrapib may constitute contributory factors In addition the LDLparticle-lowering effect of CETP inhibitors appears to become atte-nuated when a CETP inhibitor is added to a statin thus leading to dis-cordance between the measured LDL-C and apoB reductions48 Theongoing 30 000-person REVEAL trial with the CETP inhibitor anace-trapib will provide more definitive evidence of the effect of adding aCETP inhibitor to a statin on the risk of cardiovascular events

Criteria for causality

A critical appraisal of the evidence discussed in this review demon-strates that the association between plasma LDL-C concentrationand the risk of ASCVD satisfies all the criteria for causality (Table 1)49

Indeed the prospective epidemiologic studies Mendelian randomiza-tion studies and randomized intervention trials all demonstrate a re-markably consistent dose-dependent log-linear association betweenthe absolute magnitude of exposure to LDL-C and the risk ofASCVD and together demonstrate that the effect of LDL-C on therisk of ASCVD increases with increasing duration of exposure (Figure2) This concordance between multiple lines of evidence most not-ably the remarkable concordance between the unbiased naturallyrandomized genetic data and the results of numerous randomizedintervention trials using multiple different agents to reduce LDL-Cprovides overwhelming clinical evidence that LDL causes ASCVD andthat lowering LDL reduces the risk of cardiovascular events

Evidence for the cumulative effectof exposure to LDL on ASCVD

Mendelian randomization studies suggest that long-term exposure tolower LDL-C is associated with up to a three-fold greater proportionalreduction in the risk of cardiovascular disease per unit reduction in LDL-C when compared with shorter term treatment with a statin startedlater in life after atherosclerosis has developed28 This finding suggeststhat the causal effect of LDL on the risk of ASCVD is determined byboth the absolute magnitude and the cumulative duration of exposureto LDL-C This finding is consistent with increasing benefit observedover time in the long-term follow-up of the WOSCOPS trial50

Because the effect of LDL-C on the risk of ASCVD appears to beboth causal and cumulative over time lowering the plasma LDL-Clevel earlier than is currently recommended may result in propor-tionally greater reductions in the lifetime risk of ASCVD than whichis estimated in short-term randomized trials Integrating the availableevidence from Mendelian randomization studies and randomized tri-als suggests that each millimole per litre reduction in LDL-C reduces

Low-density lipoproteins cause atherosclerotic cardiovascular disease 9

Table 2 Expected proportional risk reduction based on pre-treatment low-density lipoprotein cholesterol (LDL-C)absolute magnitude of LDL-C reduction and total duration of therapy

Baseline Absolute reduction Duration of treatment exposure Guideline

LDL-C (mmolL) LDL-C (mmolL) [expected proportional risk reduction ()] recommended treatment5152

5 years 10 years 20 years 30 years 40 years

7 35 58 68 81 89 93 Yes (due to markedly elevated LDL-C)

7 30 53 62 76 85 90

7 25 46 56 70 79 86

7 20 39 48 61 71 79

7 15 31 39 51 61 69

5 25 46 56 70 79 86 Yes (due to markedly elevated LDL-C)

5 20 39 48 61 71 79

5 15 31 39 51 61 69

5 10 22 28 38 46 54

3 15 31 39 51 61 69 Depends on risk of ASCVD

3 10 22 28 38 46 54

3 05 12 15 21 27 32

2 10 22 28 38 46 54 Depends on risk of ASCVD

2 05 12 15 21 27 32

Recommendations for treatment expected clinical benefit per unit reduction in LDL-C expressed as the expected proportional risk reduction () The proportional reduction inshort-term risk is based on an expected 22 reduction in risk per millimole per litre reduction in LDL-C over 5 years as estimated from randomized trials and is calculated as[(1-078 (absolute reduction in LDL-C in mmolL)) 100] The proportional reduction in long-term risk is based on an expected 54 reduction in risk per millimole per litre reduction inLDL-C over 40 years of exposure as estimated from Mendelian randomization studies and is calculated as [(1-046 (absolute reduction in LDL-C in mmolL)) 100] The expected propor-tional risk reduction per mmolL reduction in LDL-C for any specific treatment duration is calculated as [(1-e(-0249 thorn (number of years of treatment - 5) (-00152))) 100]ASCVD atherosclerotic cardiovascular disease

Table 3 Expected short-term absolute risk reduction and number needed to treat based on baseline absolute risk ofcardiovascular disease and pre-treatment low-density lipoprotein cholesterol (LDL-C) with 5 years of treatment tolower LDL-C

10-year

absolute risk

of CVD ()

Baseline

LDL-C mmolL

(mgdL)

LDL-C after

50 reduction

mmolL (mgdL)

Proportional risk

reduction ()

10-year Absolute

risk () after 50LDL-C reduction

ARR () NNT Guideline

recommended

treatment5152

20 5 (200) 25 (100) 428 114 86 117 Yes (based on 10-

year risk of

ASCVD)

20 4 (160) 20 (80) 36 128 72 139

20 3 (120) 15 (60) 284 143 57 176

20 2 (80) 10 (40) 20 16 4 25

15 5 (200) 25 (100) 428 86 64 156 Yes (based on 10-

year risk of

ASCVD)

15 4 (160) 20 (80) 36 96 54 185

15 3 (120) 15 (60) 284 107 43 234

15 2 (80) 10 (40) 20 12 3 333

10 5 (200) 25 (100) 428 57 43 234 Yes (based on 10-

year risk of

ASCVD)

10 4 (160) 20 (80) 36 64 36 278

10 3 (120) 15 (60) 284 72 28 352

10 2 (80) 10 (40) 20 8 2 50

5 5 (200) 25 (100) 428 29 21 468 No (based on 10-

year risk of

ASCVD)

5 4 (160) 20 (80) 36 32 18 556

5 3 (120) 15 (60) 284 36 14 703

5 2 (80) 10 (40) 20 4 1 100

Recommendations for treatment expected clinical benefit per unit reduction in LDL-C The proportional reduction in short-term risk is based on an expected 22 reduction in riskper millimole per litre reduction in LDL-C over 5 years as estimated from randomized trials and is calculated as [(1-078 (absolute reduction in LDL-C in mmolL))100] The absolute risk re-duction is calculated as [ARR = Column 1 - Column 5] and the number needed to treat is calculated as [(1 ARR) 100]CVD cardiovascular disease ASCVD atherosclerotic cardiovascular disease ARR absolute risk reduction NNT number needed to treat

10 BA Ference et al

the relative risk of ASCVD events by 10 during the first year oftreatment by 16 after 2 years of treatment and by 20 after 3years of treatment presumably related to the stabilization of anyexisting underlying plaque burden33 Each subsequent year of treat-ment after the third might then be expected to result in a further15 proportional reduction in ASCVD events per millimole per litreper year (Tables 2ndash4) Therefore 5 years of treatment with a lipid-lowering agent should reduce the relative risk of ASCVD by 20ndash25 per millimole per litre reduction in LDL-C while 40 years oftreatment (or approximately equivalently 40 years of exposure tolower LDL-C) would be expected to reduce ASCVD events by50ndash55 per millimole per litre lower LDL-C (Figure 2)

Recommendations for treatment

The evidence clearly indicates that reducing LDL-C by increasing thenumbers of hepatic LDL receptors can reduce ASCVD (Figure 4)Because reducing LDL-C should have a constant effect on the risk ofASCVD per unit absolute reduction in LDL-C and because LDL-Chas both a causal and a cumulative effect on the risk of ASCVD theproportional reduction in risk that a person can expect from reduc-ing LDL-C will depend on their baseline LDL-C level the absolutemagnitude of the reduction in LDL-C and the duration of therapySimilarly the absolute risk reduction that a person can expect fromreducing LDL-C will depend on their baseline risk of ASCVD base-line LDL-C level the absolute magnitude of the reduction in LDL-Cand the duration of LDL-C-lowering therapy

Tables 2ndash4 provide estimates of the potential clinical benefit thatcan be achieved by lowering plasma LDL-C concentration based on apersonrsquos baseline risk of ASCVD baseline LDL-C and the durationof lipid-lowering therapy expressed as both expected proportionaland absolute risk reductions Physicians can use this information todiscuss with their patients about the potential benefits of loweringLDL-C based on the causal effect of LDL on the risk of ASCVD

Impact of other exposures on thecasual effect of LDL on ASCVD

In addition to LDL multiple other exposures have been reportedto be associated with the risk of ASCVD including elevated sys-tolic blood pressure diabetes and tobacco smoking1 Both theMendelian randomization studies and the meta-analyses ofrandomized trials demonstrate that changes in LDL-C have a veryconsistent proportional effect on the risk of ASCVD across differ-ent levels of various risk factors As a result a given absolute re-duction in LDL-C is associated with the same proportionalreduction in the risk of ASCVD regardless of the presence or ab-sence of other risk factors However because persons with morerisk factors have a higher absolute rate of ASCVD in comparisonwith persons with fewer risk factors the constant proportionalreduction in risk per millimole per litre reduction in LDL-C willtranslate into greater absolute risk reductions for persons withincreasing numbers of risk factors or among persons who are at ahigher risk of cardiovascular disease more generally

Table 4 Expected long-term absolute risk reduction and number needed to treat based on baseline absolute risk ofcardiovascular disease and pre-treatment low-density lipoprotein cholesterol (LDL-C) with 30 years of treatment (orexposure) to lower LDL-C

30-year

Absolute risk

of CVD ()

Baseline LDL-C

mmolL (mgdL)

LDL-C after

reduction

mmolL (mgdL)

Proportional

risk reduction ()

30-year absolute

risk () after 50LDL-C reduction

ARR() NNT Guideline

recommended

treatment5152

60 5 (200) 25 (100) 823 106 494 2 Individualized (based

on lifetime risk of

ASCVD)

60 4 (160) 20 (80) 75 15 45 22

60 3 (120) 15 (60) 646 212 389 26

60 2 (80) 10 (40) 50 30 30 33

45 5 (200) 25 (100) 823 8 37 27 Individualized (based

on lifetime risk of

ASCVD)

45 4 (160) 20 (80) 75 113 338 3

45 3 (120) 15 (60) 646 159 291 34

45 2 (80) 10 (40) 50 225 225 44

30 5 (200) 25 (100) 823 53 247 4 Individualized (based

on lifetime risk of

ASCVD)

30 4 (160) 20 (80) 75 75 225 44

30 3 (120) 15 (60) 646 106 194 52

30 2 (80) 10 (40) 50 15 15 67

15 5 (200) 25 (100) 823 27 123 81 Individualized (based

on lifetime risk of

ASCVD)

15 4 (160) 20 (80) 75 38 113 89

15 3 (120) 15 (60) 646 53 97 103

15 2 (80) 10 (40) 50 75 75 133

Recommendations for treatment expected clinical benefit per unit reduction in LDL-C The proportional reduction in long-term risk is based on an expected 54 reduction in riskper millimole per litre reduction in LDL-C over 40 years of exposure as estimated from Mendelian randomization studies and is calculated as [(1-046 (absolute reduction in LDL-C in mmol

L))100] The absolute risk reduction is calculated as [ARR = Column 1 - Column 5] and the number needed to treat is calculated as [(1ARR) 100]CVD cardiovascular disease ASCVD atherosclerotic cardiovascular disease ARR absolute risk reduction NNT number needed to treat

Low-density lipoproteins cause atherosclerotic cardiovascular disease 11

the relative risk of ASCVD events by 10 during the first year oftreatment by 16 after 2 years of treatment and by 20 after 3years of treatment presumably related to the stabilization of anyexisting underlying plaque burden33 Each subsequent year of treat-ment after the third might then be expected to result in a further15 proportional reduction in ASCVD events per millimole per litreper year (Tables 2ndash4) Therefore 5 years of treatment with a lipid-lowering agent should reduce the relative risk of ASCVD by 20ndash25 per millimole per litre reduction in LDL-C while 40 years oftreatment (or approximately equivalently 40 years of exposure tolower LDL-C) would be expected to reduce ASCVD events by50ndash55 per millimole per litre lower LDL-C (Figure 2)

Recommendations for treatment

The evidence clearly indicates that reducing LDL-C by increasing thenumbers of hepatic LDL receptors can reduce ASCVD (Figure 4)Because reducing LDL-C should have a constant effect on the risk ofASCVD per unit absolute reduction in LDL-C and because LDL-Chas both a causal and a cumulative effect on the risk of ASCVD theproportional reduction in risk that a person can expect from reduc-ing LDL-C will depend on their baseline LDL-C level the absolutemagnitude of the reduction in LDL-C and the duration of therapySimilarly the absolute risk reduction that a person can expect fromreducing LDL-C will depend on their baseline risk of ASCVD base-line LDL-C level the absolute magnitude of the reduction in LDL-Cand the duration of LDL-C-lowering therapy

Tables 2ndash4 provide estimates of the potential clinical benefit thatcan be achieved by lowering plasma LDL-C concentration based on apersonrsquos baseline risk of ASCVD baseline LDL-C and the durationof lipid-lowering therapy expressed as both expected proportionaland absolute risk reductions Physicians can use this information todiscuss with their patients about the potential benefits of loweringLDL-C based on the causal effect of LDL on the risk of ASCVD

Impact of other exposures on thecasual effect of LDL on ASCVD

In addition to LDL multiple other exposures have been reportedto be associated with the risk of ASCVD including elevated sys-tolic blood pressure diabetes and tobacco smoking1 Both theMendelian randomization studies and the meta-analyses ofrandomized trials demonstrate that changes in LDL-C have a veryconsistent proportional effect on the risk of ASCVD across differ-ent levels of various risk factors As a result a given absolute re-duction in LDL-C is associated with the same proportionalreduction in the risk of ASCVD regardless of the presence or ab-sence of other risk factors However because persons with morerisk factors have a higher absolute rate of ASCVD in comparisonwith persons with fewer risk factors the constant proportionalreduction in risk per millimole per litre reduction in LDL-C willtranslate into greater absolute risk reductions for persons withincreasing numbers of risk factors or among persons who are at ahigher risk of cardiovascular disease more generally

Table 4 Expected long-term absolute risk reduction and number needed to treat based on baseline absolute risk ofcardiovascular disease and pre-treatment low-density lipoprotein cholesterol (LDL-C) with 30 years of treatment (orexposure) to lower LDL-C

30-year

Absolute risk

of CVD ()

Baseline LDL-C

mmolL (mgdL)

LDL-C after

reduction

mmolL (mgdL)

Proportional

risk reduction ()

30-year absolute

risk () after 50LDL-C reduction

ARR() NNT Guideline

recommended

treatment5152

60 5 (200) 25 (100) 823 106 494 2 Individualized (based

on lifetime risk of

ASCVD)

60 4 (160) 20 (80) 75 15 45 22

60 3 (120) 15 (60) 646 212 389 26

60 2 (80) 10 (40) 50 30 30 33

45 5 (200) 25 (100) 823 8 37 27 Individualized (based

on lifetime risk of

ASCVD)

45 4 (160) 20 (80) 75 113 338 3

45 3 (120) 15 (60) 646 159 291 34

45 2 (80) 10 (40) 50 225 225 44

30 5 (200) 25 (100) 823 53 247 4 Individualized (based

on lifetime risk of

ASCVD)

30 4 (160) 20 (80) 75 75 225 44

30 3 (120) 15 (60) 646 106 194 52

30 2 (80) 10 (40) 50 15 15 67

15 5 (200) 25 (100) 823 27 123 81 Individualized (based

on lifetime risk of

ASCVD)

15 4 (160) 20 (80) 75 38 113 89

15 3 (120) 15 (60) 646 53 97 103

15 2 (80) 10 (40) 50 75 75 133

Recommendations for treatment expected clinical benefit per unit reduction in LDL-C The proportional reduction in long-term risk is based on an expected 54 reduction in riskper millimole per litre reduction in LDL-C over 40 years of exposure as estimated from Mendelian randomization studies and is calculated as [(1-046 (absolute reduction in LDL-C in mmol

L))100] The absolute risk reduction is calculated as [ARR = Column 1 - Column 5] and the number needed to treat is calculated as [(1ARR) 100]CVD cardiovascular disease ASCVD atherosclerotic cardiovascular disease ARR absolute risk reduction NNT number needed to treat

Low-density lipoproteins cause atherosclerotic cardiovascular disease 11

Finally an important area for future research is to identify personswho are most likely to benefit from LDL-C-lowering therapies Theprobability that LDL will be retained within the arterial intima leadingto the development and growth of atherosclerotic plaque increaseswith increasing concentration of circulating LDL particles6 Becauseretention of LDL particles is a probabilistic event one would expecttherefore that people with similar LDL-C level will have a distribu-tion of underlying atherosclerotic plaque burdens Genetic factorsmay influence whether a person is more or less likely to retain LDL-C within the arterial intima or influence the degree to which reten-tion of LDL particles triggers the inflammatory process or oxidativechanges that influence the rate of plaque growth and the propensityfor plaque disruption

Indeed many of the genetic variants most strongly associated withASCVD in GWAS are in genes that encode for components of thearterial wall which may modify susceptibility to retaining LDL ormodify responses to the accumulated LDL within the artery wall27

This hypothesis is consistent with the results of a meta-analysis of sta-tin trials that reported that persons with the highest tertile of a gen-etic ASCVD risk score derived a greater proportional reduction inthe risk of cardiovascular events per millimole per litre reduction inLDL-C than did persons with the lowest tertile of the genetic riskscore53 The genetic score used in this study may have included vari-ants that lead to a greater susceptibility to retain LDL particlesIdentifying genetic and other factors that influence the likelihood ofretaining LDL particles within the intima is an active area of ongoingresearch and may eventually help to personalize the prevention ofcardiovascular disease by identifying persons who are most vulner-able to the deleterious effects of LDL-C and therefore are most likelyto benefit from therapies that lower LDL-C

Conclusions

Considered together the strong and consistent evidence from thegenetic studies prospective epidemiologic cohort studies Mendelianrandomization studies and randomized intervention trials discussedhere supported by mechanistic evidence to be presented in the se-cond Consensus Statement on LDL causality establishes that LDL isnot merely a biomarker of increased risk but a causal factor in the

pathophysiology of ASCVD The key implications for this conclusionare presented in Box 1

FundingTravel and meeting logistics were supported by unrestricted educationalgrants from MSD and Amgen to the European Atherosclerosis SocietyThese companies were not present at the Consensus Panel meetingshad no role in the design or content of the manuscript and had no rightto approve or disapprove the final document Funding to pay the OpenAccess publication charges for this article was provided by the EuropeanAtherosclerosis Society

Conflict of interest JB has received research grants from AmgenAstraZeneca NovoNordisk Pfizer and RegeneronSanofi and honorariafor consultancy and lectures from Amgen AstraZeneca Eli Lilly MerckNovo-Nordisk Pfizer and RegeneronSanofi EB has received honorariafrom AstraZeneca Amgen Genfit MSD Sanofi-Regeneron UnileverDanone Aegerion Chiesi Rottapharm Lilly and research grants fromAmgen Danone and Aegerion ALC has received research grants to hisinstitution from Amgen Astra-Zeneca Merck RegeneronSanofi andSigma Tau and honoraria for advisory boards consultancy or speakerbureau from Abbot Aegerion Amgen AstraZeneca Eli Lilly GenzymeMerckMSDMylan Pfizer Rottapharm and Sanofi-Regeneron MJC hasreceived research grants from MSD Kowa Pfizer and Randox and hono-raria for consultancyspeaker activities from Amgen Kowa Merck SanofiServier Unilever and Regeneron SF has the following disclosures forthe last 12 months Compensated consultant and advisory activities withMerck Kowa Sanofi Amgen Amarin and Aegerion BAF has receivedresearch grants from Merck Amgen and Esperion Therapeutics andreceived honoraria for lectures consulting andor advisory board mem-bership from Merck Amgen Esperion Ionis and the American Collegeof Cardiology IG has received speaker fees from MSD and Pfizer relatingto cardiovascular risk estimation and lipid guidelines and consultancyspeaker fee from Amgen HNG has received research grants fromMerck Sanofi-Regeneron and Amgen He consults for Merck SanofiRegeneron Lilly Kowa Resverlogix Boehringer Ingelheim RAH hasreceived research grants from Aegerion Amgen The MedicinesCompany Pfizer and Sanofi He consults for Amgen Aegerion BostonHeart Diagnostics Gemphire Lilly and Sanofi JDH reports honorariaresearch grants from Aegerion Alnylam Catabasis Lilly Merck PfizerNovartis Regeneron Sanofi RMK is a Member Merck GlobalAtherosclerosis Advisory Board UL has received honoraria for lecturesandor consulting from Amgen Medicines Company Astra ZenecaMSD Berlin Chemie Bayer Abbott and Sanofi UL aufs has receivedhonoraria for board membership consultancy and lectures from AmgenMSD Sanofi and Servier LM has received honoraria for consultancy andlectures from Amgen Danone Kowa Merck and Sanofi-RegeneronSJN has received research support from Amgen AstraZeneca AntheraCerenis Novartis Eli Lilly Esperion Resverlogix Sanofi-RegeneronInfraReDx and LipoScience and is a consultant for Amgen AstraZenecaBoehringer Ingelheim CSL Behring Eli Lilly Merck Takeda Pfizer RocheSanofi-Regeneron Kowa and Novartis BGN reports consultancies andhonoraria for lectures from AstraZeneca Sanofi Regeneron AegerionFresenius B Braun Kaneka Amgen CJP has received research supportfrom Roche MSD and honoraria from MSD SanofiRegeneron Amgenand Pfizer FJR has received grantsresearch support from Amgen andSanofi and has received speaker fees or honoraria for consultation fromAstraZeneca Merck Amgen and Sanofi KKR has received researchgrants from Amgen Sanofi-Regeneron and Pfizer and honoraria for lec-tures advisory boards or as a steering committee member fromAegerion Amgen Sanofi-Regeneron Pfizer AstraZeneca Cerenis ISISPharma Medco Resverlogix Kowa Novartis Cipla Lilly Algorithm

Box 1 Low-density lipoprotein (LDL) as a causalfactor for atherosclerotic cardiovascular diseasekey implications

bull Cumulative LDL arterial burden is a central determinant for the

initiation and progression of atherosclerotic cardiovascular diseasebull The lower the LDL cholesterol (LDL-C) level attained by agents

that primarily target LDL receptors the greater the clinical benefit

accruedbull Both proportional (relative) risk reduction and absolute risk reduc-

tion relate to the magnitude of LDL-C reductionbull Lowering LDL-C in individuals at high cardiovascular risk earlier ra-

ther than later appears advisable especially in those with familial

hypercholesterolaemia

12 BA Ference et al

Takeda Boehringer Ingelheim MSD Esperion and AbbieVie HS hasreceived research grants from AstraZeneca MSD Bayer Vital sanofi-aventis and Pfizer and honoraria for speaker fees from AstraZenecaMSD Genzyme sanofi-aventis and Synlab He has consulted for MSDand AstraZeneca MRT has received speaker fees from Amgen AstraZeneca Chiesi Pharma and Eli Lilly and speaker fees and research supportfrom Amgen Sanofi Aventis and Novo Nordisk She has consulted forAstraZeneca LT has received research funding andor honoraria for ad-visory boards consultancy or speaker bureau from Abbott MylanActelion Aegerion Amgen AstraZeneca Bayer Boehringer IngelheimDaiichi-Sankyo GlaxoSmithKline Menarini Merck Novartis PfizerSanofi-Regeneron Servier and Synageva GFW has received researchsupport from Amgen and Sanofi-Regeneron OW has received honora-ria for lectures or consultancy from Sanofi Amgen MSD and Astra-Zeneca BvS and JKS report no disclosures

References1 Yusuf S Hawken S Ounpuu S Dans T Avezum A Lanas F McQueen M Budaj

A Pais P Varigos J Lisheng L INTERHEART Study Investigators Effect of poten-tially modifiable risk factors associated with myocardial infarction in 52 countries(the INTERHEART study) case-control study Lancet 2004364937ndash952

2 Goldstein JL Brown MS A century of cholesterol and coronaries from plaquesto genes to statins Cell 2015161161ndash172

3 DuBroff R Cholesterol paradox a correlate does not a surrogate make EvidBased Med 20172215ndash19

4 Ravnskov U Diamond DM Hama R Hamazaki T Hammarskjold B Hynes NKendrick M Langsjoen PH Malhotra A Mascitelli L McCully KS Ogushi YOkuyama H Rosch PJ Schersten T Sultan S Sundberg R Lack of an associationor an inverse association between low-density-lipoprotein cholesterol and mor-tality in the elderly a systematic review BMJ Open 20166e010401

5 Camejo G Lopez A Vegas H Paoli H The participation of aortic proteins in theformation of complexes between low density lipoproteins and intima-media ex-tracts Atherosclerosis 19752177ndash91

6 Skalen K Gustafsson M Rydberg EK Hulten LM Wiklund O Innerarity TLBoren J Subendothelial retention of atherogenic lipoproteins in early athero-sclerosis Nature 2002417750ndash754

7 Schwenke DC Carew TE Initiation of atherosclerotic lesions in cholesterol-fedrabbits II Selective retention of LDL vs selective increases in LDL permeabilityin susceptible sites of arteries Arteriosclerosis 19899908ndash918

8 Frank JS Fogelman AM Ultrastructure of the intima in WHHL and cholesterol-fed rabbit aortas prepared by ultra-rapid freezing and freeze-etching J Lipid Res198930967ndash978

9 Tabas I Williams KJ Boren J Subendothelial lipoprotein retention as the initiatingprocess in atherosclerosis update and therapeutic implications Circulation20071161832ndash1844

10 Nordestgaard BG Zilversmit DB Large lipoproteins are excluded from the ar-terial wall in diabetic cholesterol-fed rabbits J Lipid Res 1988291491ndash1500

11 Descamps OS Bruniaux M Guilmot PF Tonglet R Heller FR Lipoprotein con-centrations in newborns are associated with allelic variations in their mothersAtherosclerosis 2004172287ndash298

12 Chapman MJ Animal lipoproteins chemistry structure and comparative aspectsJ Lipid Res 198021789ndash853

13 Nordestgaard BG Chapman MJ Humphries SE Ginsberg HN Masana LDescamps OS Wiklund O Hegele RA Raal FJ Defesche JC Wiegman A SantosRD Watts GF Parhofer KG Hovingh GK Kovanen PT Boileau C Averna MBoren J Bruckert E Catapano AL Kuivenhoven JA Pajukanta P Ray KStalenhoef AF Stroes E Taskinen MR Tybjaeligrg-Hansen A EuropeanAtherosclerosis Society Consensus Panel Familial hypercholesterolaemia isunderdiagnosed and undertreated in the general population guidance for clin-icians to prevent coronary heart disease Consensus Statement of the EuropeanAtherosclerosis Society Eur Heart J 2013343478ndash3490

14 Cuchel M Bruckert E Ginsberg HN Raal FJ Santos RD Hegele RAKuivenhoven JA Nordestgaard BG Descamps OS Steinhagen-Thiessen ETybjaeligrg-Hansen A Watts GF Averna M Boileau C Boren J Catapano ALDefesche JC Hovingh GK Humphries SE Kovanen PT Masana L Pajukanta PParhofer KG Ray KK Stalenhoef AF Stroes E Taskinen MR Wiegman AWiklund O Chapman MJ European Atherosclerosis Society Consensus Panel onFamilial Hypercholesterolaemia Homozygous familial hypercholesterolaemianew insights and guidance for clinicians to improve detection and clinical man-agement A position paper from the Consensus Panel on FamilialHypercholesterolaemia of the European Atherosclerosis Society Eur Heart J2014352146ndash2157

15 Khera AV Won HH Peloso GM Lawson KS Bartz TM Deng X van LeeuwenEM Natarajan P Emdin CA Bick AG Morrison AC Brody JA Gupta N NomuraA Kessler T Duga S Bis JC van Duijn CM Cupples LA Psaty B Rader DJ DaneshJ Schunkert H McPherson R Farrall M Watkins H Lander E Wilson JG CorreaA Boerwinkle E Merlini PA Ardissino D Saleheen D Gabriel S Kathiresan SDiagnostic yield of sequencing familial hypercholesterolemia genes in patients withsevere hypercholesterolemia J Am Coll Cardiol 2016672578ndash2589

16 Benn M Watts GF Tybjaeligrg-Hansen A Nordestgaard BG Mutations causative offamilial hypercholesterolaemia screening of 98 098 individuals from theCopenhagen General Population Study estimated a prevalence of 1 in 217 EurHeart J 2016371384ndash1394

17 Raal FJ Pilcher GJ Waisberg R Buthelezi EP Veller MG Joffe BI Low-densitylipoprotein cholesterol bulk is the pivotal determinant of atherosclerosis in famil-ial hypercholesterolemia Am J Cardiol 1999831330ndash1333

18 Schmidt HH Hill S Makariou EV Feuerstein IM Dugi KA Hoeg JM Relationship ofcholesterol-year score to severity of calcific atherosclerosis and tissue depositionin homozygous familial hypercholesterolemia Am J Cardiol 199677575ndash580

19 Wiegman A Gidding SS Watts GF Chapman MJ Ginsberg HN Cuchel M Ose LAverna M Boileau C Boren J Bruckert E Catapano AL Defesche JC DescampsOS Hegele RA Hovingh GK Humphries SE Kovanen PT Kuivenhoven JA MasanaL Nordestgaard BG Pajukanta P Parhofer KG Raal FJ Ray KK Santos RDStalenhoef AF Steinhagen Thiessen E Stroes ES Taskinen MR Tybjaeligrg-Hansen AWiklund O European Atherosclerosis Society Consensus Panel Familial hyper-cholesterolaemia in children and adolescents gaining decades of life by optimizingdetection and treatment Eur Heart J 2015362425ndash2437

20 Cohen JC Boerwinkle E Mosley TH Jr Hobbs HH Sequence variations inPCSK9 low LDL and protection against coronary heart disease N Engl J Med20063541264ndash1272

21 Emerging Risk Factors C Di Angelantonio E Gao P Pennells L Kaptoge SCaslake M Thompson A Butterworth AS Sarwar N Wormser D Saleheen DBallantyne CM Psaty BM Sundstrom J Ridker PM Nagel D Gillum RF Ford IDucimetiere P Kiechl S Koenig W Dullaart RP Assmann G Drsquoagostino RB SrDagenais GR Cooper JA Kromhout D Onat A Tipping RW Gomez-de-la-Camara A Rosengren A Sutherland SE Gallacher J Fowkes FG Casiglia EHofman A Salomaa V Barrett-Connor E Clarke R Brunner E Jukema JWSimons LA Sandhu M Wareham NJ Khaw KT Kauhanen J Salonen JT HowardWJ Nordestgaard BG Wood AM Thompson SG Boekholdt SM Sattar NPackard C Gudnason V Danesh J Lipid-related markers and cardiovascular dis-ease prediction JAMA 20123072499ndash2506

22 Prospective Studies Collaboration Lewington S Whitlock G Clarke R SherlikerP Emberson J Halsey J Qizilbash N Peto R Collins R Blood cholesterol andvascular mortality by age sex and blood pressure a meta-analysis of individ-ual data from 61 prospective studies with 55000 vascular deaths Lancet20073701829ndash1839

23 Lawlor DA Harbord RM Sterne JA Timpson N Davey Smith G Mendelian ran-domization using genes as instruments for making causal inferences in epidemi-ology Stat Med 2008271133ndash1163

24 Teslovich TM Musunuru K Smith AV Edmondson AC Stylianou IM Koseki MPirruccello JP Ripatti S Chasman DI Willer CJ Johansen CT Fouchier SWIsaacs A Peloso GM Barbalic M Ricketts SL Bis JC Aulchenko YS ThorleifssonG Feitosa MF Chambers J Orho-Melander M Melander O Johnson T Li XGuo X Li M Shin Cho Y Jin Go M Jin Kim Y Lee JY Park T Kim K Sim XTwee-Hee Ong R Croteau-Chonka DC Lange LA Smith JD Song K Hua ZhaoJ Yuan X Luan J Lamina C Ziegler A Zhang W Zee RY Wright AF WittemanJC Wilson JF Willemsen G Wichmann HE Whitfield JB Waterworth DMWareham NJ Waeber G Vollenweider P Voight BF Vitart V Uitterlinden AGUda M Tuomilehto J Thompson JR Tanaka T Surakka I Stringham HM SpectorTD Soranzo N Smit JH Sinisalo J Silander K Sijbrands EJ Scuteri A Scott JSchlessinger D Sanna S Salomaa V Saharinen J Sabatti C Ruokonen A Rudan IRose LM Roberts R Rieder M Psaty BM Pramstaller PP Pichler I Perola MPenninx BW Pedersen NL Pattaro C Parker AN Pare G Oostra BAOrsquodonnell CJ Nieminen MS Nickerson DA Montgomery GW Meitinger TMcPherson R McCarthy MI McArdle W Masson D Martin NG Marroni FMangino M Magnusson PK Lucas G Luben R Loos RJ Lokki ML Lettre GLangenberg C Launer LJ Lakatta EG Laaksonen R Kyvik KO Kronenberg FKonig IR Khaw KT Kaprio J Kaplan LM Johansson A Jarvelin MR Janssens ACIngelsson E Igl W Kees Hovingh G Hottenga JJ Hofman A Hicks AAHengstenberg C Heid IM Hayward C Havulinna AS Hastie ND Harris TBHaritunians T Hall AS Gyllensten U Guiducci C Groop LC Gonzalez E GiegerC Freimer NB Ferrucci L Erdmann J Elliott P Ejebe KG Doring A DominiczakAF Demissie S Deloukas P de Geus EJ de Faire U Crawford G Collins FSChen YD Caulfield MJ Campbell H Burtt NP Bonnycastle LL Boomsma DIBoekholdt SM Bergman RN Barroso I Bandinelli S Ballantyne CM Assimes TLQuertermous T Altshuler D Seielstad M Wong TY Tai ES Feranil AB KuzawaCW Adair LS Taylor HA Jr Borecki IB Gabriel SB Wilson JG Holm HThorsteinsdottir U Gudnason V Krauss RM Mohlke KL Ordovas JM Munroe

Low-density lipoproteins cause atherosclerotic cardiovascular disease 13

PB Kooner JS Tall AR Hegele RA Kastelein JJ Schadt EE Rotter JI BoerwinkleE Strachan DP Mooser V Stefansson K Reilly MP Samani NJ Schunkert HCupples LA Sandhu MS Ridker PM Rader DJ van Duijn CM Peltonen LAbecasis GR Boehnke M Kathiresan S Biological clinical and population rele-vance of 95 loci for blood lipids Nature 2010466707ndash713

25 Global Lipids Genetics Consortium Discovery and refinement of loci associatedwith lipid levels Nat Genet 2013451274ndash1283

26 Ference BA Mendelian randomization studies using naturally randomized gen-etic data to fill evidence gaps Curr Opin Lipidol 201526566ndash571

27 CARDIoGRAMplusC4D Consortium A comprehensive 1000 Genomes-basedgenome-wide association meta-analysis of coronary artery disease Nat Genet2015471121ndash1130

28 Ference BA Yoo W Alesh I Mahajan N Mirowska KK Mewada A Kahn J AfonsoL Williams KA Sr Flack JM Effect of long-term exposure to lower low-densitylipoprotein cholesterol beginning early in life on the risk of coronary heart diseasea Mendelian randomization analysis J Am Coll Cardiol 2012602631ndash2639

29 Linsel-Nitschke P Gotz A Erdmann J Braenne I Braund P Hengstenberg CStark K Fischer M Schreiber S El Mokhtari NE Schaefer A Schrezenmeier JRubin D Hinney A Reinehr T Roth C Ortlepp J Hanrath P Hall AS ManginoM Lieb W Lamina C Heid IM Doering A Gieger C Peters A Meitinger TWichmann HE Konig IR Ziegler A Kronenberg F Samani NJ Schunkert HLifelong reduction of LDL-cholesterol related to a common variant in the LDL-receptor gene decreases the risk of coronary artery disease a MendelianRandomisation study PLoS One 20083e2986

30 Holmes MV Asselbergs FW Palmer TM Drenos F Lanktree MB Nelson CP DaleCE Padmanabhan S Finan C Swerdlow DI Tragante V van Iperen EPSivapalaratnam S Shah S Elbers CC Shah T Engmann J Giambartolomei C White JZabaneh D Sofat R McLachlan S UCLEB Consortium Doevendans PA BalmforthAJ Hall AS North KE Almoguera B Hoogeveen RC Cushman M Fornage M PatelSR Redline S Siscovick DS Tsai MY Karczewski KJ Hofker MH Verschuren WMBots ML van der Schouw YT Melander O Dominiczak AF Morris R Ben-ShlomoY Price J Kumari M Baumert J Peters A Thorand B Koenig W Gaunt TRHumphries SE Clarke R Watkins H Farrall M Wilson JG Rich SS de Bakker PILange LA Davey Smith G Reiner AP Talmud PJ Kivimeuroaki M Lawlor DA DudbridgeF Samani NJ Keating BJ Hingorani AD Casas JP Mendelian randomization of bloodlipids for coronary heart disease Eur Heart J 201536539ndash550

31 Ference BA Majeed F Penumetcha R Flack JM Brook RD Effect of naturally ran-dom allocation to lower low-density lipoprotein cholesterol on the risk of coron-ary heart disease mediated by polymorphisms in NPC1L1 HMGCR or both a 2 x2 factorial Mendelian randomization study J Am Coll Cardiol 2015651552ndash1261

32 Cholesterol Treatment Trialistsrsquo (CTT) Collaboration Baigent C Blackwell LEmberson J Holland LE Reith C Bhala N Peto R Barnes EH Keech A Simes JCollins R Efficacy and safety of more intensive lowering of LDL cholesterol ameta-analysis of data from 170 000 participants in 26 randomised trials Lancet20103761670ndash1681

33 Collins R Reith C Emberson J Armitage J Baigent C Blackwell L Blumenthal RDanesh J Smith GD5 DeMets D Evans S Law M MacMahon S Martin S NealB Poulter N Preiss D Ridker P Roberts I Rodgers A Sandercock P Schulz KSever P Simes J Smeeth L Wald N Yusuf S Peto R Interpretation of the evi-dence for the efficacy and safety of statin therapy Lancet 20163882532ndash2561

34 Boekholdt SM Arsenault BJ Mora S Pedersen TR LaRosa JC Nestel PJ SimesRJ Durrington P Hitman GA Welch KM DeMicco DA Zwinderman AHClearfield MB Downs JR Tonkin AM Colhoun HM Gotto AM Jr Ridker PMKastelein JJ Association of LDL cholesterol non-HDL cholesterol and apolipo-protein B levels with risk of cardiovascular events among patients treated withstatins a meta-analysis JAMA 20123071302ndash1309

35 Nissen SE Nicholls SJ Sipahi I Libby P Raichlen JS Ballantyne CM Davignon JErbel R Fruchart JC Tardif JC Schoenhagen P Crowe T Cain V Wolski KGoormastic M Tuzcu EM ASTEROID Investigators Effect of very high-intensitystatin therapy on regression of coronary atherosclerosis the ASTEROID trialJAMA 20062951556ndash1565

36 Nicholls SJ Ballantyne CM Barter PJ Chapman MJ Erbel RM Libby P RaichlenJS Uno K Borgman M Wolski K Nissen SE Effect of two intensive statin regi-mens on progression of coronary disease N Engl J Med 20113652078ndash2087

37 Garcia-Calvo M Lisnock J Bull HG Hawes BE Burnett DA Braun MP Crona JHDavis HR Jr Dean DC Detmers PA Graziano MP Hughes M Macintyre DEOgawa A Orsquoneill KA Iyer SP Shevell DE Smith MM Tang YS Makarewicz AMUjjainwalla F Altmann SW Chapman KT Thornberry NA The target of ezetimibeis Niemann-Pick C1-Like 1 (NPC1L1) Proc Natl Acad Sci USA 20051028132ndash8137

38 Cannon CP Blazing MA Giugliano RP McCagg A White JA Theroux P DariusH Lewis BS Ophuis TO Jukema JW De Ferrari GM Ruzyllo W De Lucca P ImK Bohula EA Reist C Wiviott SD Tershakovec AM Musliner TA Braunwald ECaliff RM IMPROVE-IT Investigators Ezetimibe added to statin therapy afteracute coronary syndromes N Engl J Med 20153722387ndash2397

39 Baigent C Landray MJ Reith C Emberson J Wheeler DC Tomson C WannerC Krane V Cass A Craig J Neal B Jiang L Hooi LS Levin A Agodoa L Gaziano

M Kasiske B Walker R Massy ZA Feldt-Rasmussen B Krairittichai UOphascharoensuk V Fellstrom B Holdaas H Tesar V Wiecek A Grobbee Dde Zeeuw D Gronhagen-Riska C Dasgupta T Lewis D Herrington W MafhamM Majoni W Wallendszus K Grimm R Pedersen T Tobert J Armitage J BaxterA Bray C Chen Y Chen Z Hill M Knott C Parish S Simpson D Sleight PYoung A Collins R SHARP Investigators The effects of lowering LDL choles-terol with simvastatin plus ezetimibe in patients with chronic kidney disease(Study of Heart and Renal Protection) a randomised placebo-controlled trialLancet 20113772181ndash2192

40 Lauridsen BK Stender S Frikke-Schmidt R Nordestgaard BG Tybjaeligrg-HansenA Genetic variation in the cholesterol transporter NPC1L1 ischemic vasculardisease and gallstone disease Eur Heart J 2015361601ndash1608

41 Sabatine MS Giugliano RP Keech A Honarpour N Wang H Liu T WassermanSM Scott R Sever PS Pedersen TR Rationale and design of the Further cardio-vascular OUtcomes Research with PCSK9 Inhibition in subjects with ElevatedRisk (FOURIER) Trial Am Heart J 201617394ndash101

42 Sabatine MS Giugliano RP Keech AC Honarpour N Wiviott SD Murphy SAKuder JF Wang H Liu T Wasserman SM Sever PS Pedersen TR FOURIERSteering Committee and Investigators Evolocumab and Clinical Outcomes inPatients with Cardiovascular Disease N Engl J Med 2017 doi 101056NEJMoa1615664

43 Ference BA Robinson JG Brook RD Catapano AL Chapman MJ Neff DRVoros S Giugliano RP Davey Smith G Fazio S Sabatine MS Variation in PCSK9and HMGCR and risk of cardiovascular disease and diabetes N Engl J Med 20163752144ndash2153

44 Nicholls SJ Puri R Anderson T Ballantyne CM Cho L Kastelein JJ Koenig WSomaratne R Kassahun H Yang J Wasserman SM Scott R Ungi I Podolec JOphuis AO Cornel JH Borgman M Brennan DM Nissen SE Effect of evolocu-mab on progression of coronary disease in statin-treated patients TheGLAGOV randomized clinical trial JAMA 20163162373ndash2384

45 Lipid Research Clinics Program The Lipid Research Clinics coronary primaryprevention trial results reduction in the incidence of coronary artery diseaseJAMA 1984251351ndash364

46 Buchwald H Varco RL Matts JP Long JM Fitch LL Campbell GS Pearce MB YellinAE Edmiston A Smink RD Sawin HS Campos CT Hansen BJ Tuna N Karnegis JNSanmarco ME Amplatz K Castaneda-Zuniga WR Hunter DW Bissett JK Weber FJStevenson JW Leon AS Chalmers TC The POSCH Group Effect of partial ileal by-pass surgery on mortality and morbidity from coronary heart disease in patientswith hypercholesterolemia Report of the Program on the Surgical Control of theHyperlipidemias (POSCH) N Engl J Med 1990 323946ndash955

47 Silverman MG Ference BA Im K Wiviott SD Giugliano RP Grundy SMBraunwald E Sabatine MS Association between lowering LDL-C and cardiovas-cular risk reduction among different therapeutic interventions a systematic re-view and meta-analysis JAMA 20163161289ndash1297

48 Lincoff AM for the ACCELERATE Trial Investigators Lipid levels and cardiovas-cular outcome with cholesteryl ester transfer protein (CETP) inhibition by evac-etrapib in the ACCELERATE Trial Late Breaking Clinical Trials ESC Congress2016 Rome httpcongress365escardioorgSearch-Resultsvgnextkeyword=evacetrapibampYears=C365YEAR2016WCm9LoXXJjo (8 February 2017)

49 Hill AB The environment and disease Association or causation Proc R Soc Med196558295ndash300

50 Ford I Murray H McCowan C Packard CJ Long-term safety and efficacy of lower-ing low-density lipoprotein cholesterol with statin therapy 20-year follow-up ofWest of Scotland Coronary Prevention Study Circulation 20161331073ndash1080

51 Catapano AL Graham I De Backer G Wiklund O Chapman MJ Drexel HHoes AW Jennings CS Landmesser U Pedersen TR Reiner Z Riccardi GTaskinen MR Tokgozoglu L Verschuren WM Vlachopoulos C Wood DAZamorano JL on behalf of AuthorsTask Force Members 2016 ESCEAS guide-lines for the management of dyslipidaemias Eur Heart J 2016372999ndash3058

52 Stone NJ Robinson JG Lichtenstein AH Bairey Merz CN Blum CB Eckel RHGoldberg AC Gordon D Levy D Lloyd-Jones DM McBride P Schwartz JSShero ST Smith SC Jr Watson K Wilson PW Eddleman KM Jarrett NMLaBresh K Nevo L Wnek J Anderson JL Halperin JL Albert NM Bozkurt BBrindis RG Curtis LH DeMets D Hochman JS Kovacs RJ Ohman EM PresslerSJ Sellke FW Shen WK Smith SC Jr Tomaselli GF American College ofCardiologyAmerican Heart Association Task Force on Practice Guidelines 2013ACCAHA guideline on the treatment of blood cholesterol to reduce athero-sclerotic cardiovascular risk in adults a report of the American College ofCardiologyAmerican Heart Association Task Force on practice guidelinesCirculation 2014129(25 Suppl 2)S1ndashS45

53 Mega JL Stitziel NO Smith JG Chasman DI Caulfield MJ Devlin JJ Nordio FHyde CL Cannon CP Sacks FM Poulter NR Sever PS Ridker PM Braunwald EMelander O Kathiresan S Sabatine MS Genetic risk coronary heart diseaseevents and the clinical benefit of statin therapy an analysis of primary and sec-ondary prevention trials Lancet 20153852264ndash2271

14 BA Ference et al